Image to text (OCR)

Extracts text from images. Results vary with image quality and language, so check the output.

Processed in your browser. We don't receive or store your files.

JPG, PNG, WebP, BMP or GIF, up to 30 MB. Clear, straight, well-lit text works best.

Language of the text

The text engine (about 4 MB) and each language file download from this site the first time you read an image. Your image stays in your browser.

How to get text from an image

  1. Choose a photo, scan or screenshot.
  2. Tick the language or languages in the image.
  3. Select Read text. The first time, the text engine and language files download from this site.
  4. Check the text, fix any mistakes, then copy or download it.

Questions

How good are the results?

Extracts text from images. Results vary with image quality and language, so check the output. Sharp, straight, printed text gives the best results. Handwriting, low light, small text and busy backgrounds give poorer results.

Which languages can it read?

English, French, Hindi and Gujarati. Tick every language that appears in the image.

Why is the first run slower?

The text engine and language files download once from this site. After that, your browser may keep them for a while.

Can it read PDFs?

Not directly. Take a screenshot of the page, or save it as an image, and choose that.

Is my image uploaded?

No. Images are processed in your browser. We don't receive or store them.

Related tools

Made by Bumbibee Labs, a company based in Ontario, Canada.